Geek Answers Handbook

The best database for storing statistics from Erlang

I need to select a database for storing statistics (actually this is a series of data with a timestamp). I understand that almost any database can handle this, but there are several requirements:

  • he must be fast;
  • he should be able to process MUCH data (10 s gigabytes) and splicing quickly;
  • It should have a stable, supported and user-friendly interface for Erlang;
  • It must be accessible from Python;
  • it should be able to do something like a thing called "capped collections" in mongodb: a collection with a limited size, with old data being overwritten if the size reaches the limit.

I thought about mongo, but emongo seems a little dead - the last fight was made 7 months ago.

+5
source share
4 answers

Riak might be a good choice (here Comparison of Riak with MongoDB ). It is written in Erlang, distributed, fault tolerant and linearly scalable. It has clients for Erlang, Javascript, Java, PHP, Python, Ruby. REST interface, protobuf interface and so many other useful properties (Map Reduce, links, replication, preliminary / post commits, ...). It works with open source and is created using Basho . Basho’s Riak’s offer is also with some additional features (for example, multisite replication, SNMP monitoring, etc.), but the OS has a great version.

. , , Redis, Riak . Redis + Riak - .

+3

, postgresql pgsql .

+1

, , . , .

+1

redis - .

, , , ( ), .

Antirez, , , diskstore, . , .

, redis . LTRIM .

0

More articles:


All Articles